{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,16]],"date-time":"2026-05-16T00:05:25Z","timestamp":1778889925461,"version":"3.51.4"},"reference-count":48,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"9","license":[{"start":{"date-parts":[[2025,9,1]],"date-time":"2025-09-01T00:00:00Z","timestamp":1756684800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2025,9,1]],"date-time":"2025-09-01T00:00:00Z","timestamp":1756684800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,9,1]],"date-time":"2025-09-01T00:00:00Z","timestamp":1756684800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"name":"National Key R&D Program of China","award":["2023YFB3002702"],"award-info":[{"award-number":["2023YFB3002702"]}]},{"name":"Key Program of National Natural Science Foundation of China","award":["U21A20461"],"award-info":[{"award-number":["U21A20461"]}]},{"name":"Key Program of National Natural Science Foundation of China","award":["92055213"],"award-info":[{"award-number":["92055213"]}]},{"name":"Key Program of National Natural Science Foundation of China","award":["62227808"],"award-info":[{"award-number":["62227808"]}]},{"DOI":"10.13039\/501100004761","name":"Natural Science Foundation of Hainan Province","doi-asserted-by":"publisher","award":["2023GK2002"],"award-info":[{"award-number":["2023GK2002"]}],"id":[{"id":"10.13039\/501100004761","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100004761","name":"Natural Science Foundation of Hainan Province","doi-asserted-by":"publisher","award":["2024JJ2026"],"award-info":[{"award-number":["2024JJ2026"]}],"id":[{"id":"10.13039\/501100004761","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100005230","name":"Natural Science Foundation of Chongqing","doi-asserted-by":"publisher","award":["CSTB2022NSCQ-MSX1213"],"award-info":[{"award-number":["CSTB2022NSCQ-MSX1213"]}],"id":[{"id":"10.13039\/501100005230","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Parallel Distrib. Syst."],"published-print":{"date-parts":[[2025,9]]},"DOI":"10.1109\/tpds.2025.3587673","type":"journal-article","created":{"date-parts":[[2025,7,10]],"date-time":"2025-07-10T17:49:39Z","timestamp":1752169779000},"page":"1985-1997","source":"Crossref","is-referenced-by-count":3,"title":["High Performance OpenCL-Based GEMM Kernel Auto-Tuned by Bayesian Optimization"],"prefix":"10.1109","volume":"36","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-3329-0924","authenticated-orcid":false,"given":"Shengle","family":"Lin","sequence":"first","affiliation":[{"name":"College of Computer Science and Electronic Engineering, Hunan University, Changsha, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-5008-4829","authenticated-orcid":false,"given":"Guoqing","family":"Xiao","sequence":"additional","affiliation":[{"name":"College of Computer Science and Electronic Engineering, Hunan University, Changsha, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0086-6301","authenticated-orcid":false,"given":"Haotian","family":"Wang","sequence":"additional","affiliation":[{"name":"College of Computer Science and Electronic Engineering, Hunan University, Changsha, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2681-7898","authenticated-orcid":false,"given":"Wangdong","family":"Yang","sequence":"additional","affiliation":[{"name":"College of Computer Science and Electronic Engineering, Hunan University, Changsha, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2635-7716","authenticated-orcid":false,"given":"Kenli","family":"Li","sequence":"additional","affiliation":[{"name":"College of Computer Science and Electronic Engineering, Hunan University, Changsha, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-5224-4048","authenticated-orcid":false,"given":"Keqin","family":"Li","sequence":"additional","affiliation":[{"name":"College of Computer Science and Electronic Engineering, Hunan University, Hunan, China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1145\/292395.292412"},{"key":"ref2","article-title":"Gotoblas","author":"Goto","year":"2007"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1145\/2764454"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1145\/2503210.2503219"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-06486-4_7"},{"key":"ref6","article-title":"cuBLAS: A GPU-accelerated library for linear algebra operations","author":"Corporation","year":"2023"},{"key":"ref7","article-title":"rocBLAS documentation","year":"2023"},{"key":"ref8","article-title":"clblas repository","year":"2017"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1145\/3204919.3204924"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1109\/MCSoC.2015.10"},{"key":"ref11","volume-title":"Heterogeneous Computing With OpenCL 2.0","author":"Kaeli","year":"2015"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1145\/2304576.2304623"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1145\/2304576.2304624"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1145\/3577193.3593720"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1109\/TPDS.2011.311"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1109\/SC.Companion.2012.59"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1109\/ASPDAC.2013.6509639"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1145\/2830772.2830813"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1145\/3368826.3377918"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1109\/IPDPS53621.2022.00089"},{"key":"ref21","article-title":"A tutorial on Bayesian optimization of expensive cost functions, with application to active user modeling and hierarchical reinforcement learning","author":"Brochu","year":"2010"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1109\/JPROC.2015.2494218"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1007\/BF01099263"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.7551\/mitpress\/3206.001.0001"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1023\/A:1008306431147"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1145\/2664666.2664674"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1016\/j.parco.2011.10.002"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1145\/3293883.3295734"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1016\/j.jpdc.2023.104741"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1109\/H2RC51942.2020.00007"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1145\/3392717.3392757"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1109\/ISCAS45731.2020.9181281"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1145\/3503540"},{"key":"ref34","volume-title":"Dynamic Programming and Modern Control Theory","volume":"81","author":"Bellman","year":"1965"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1007\/3-540-44864-0_78"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-642-21434-9_3"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1145\/2628071.2628092"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1109\/TEVC.2019.2921598"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.1109\/Cluster48925.2021.00037"},{"key":"ref40","doi-asserted-by":"publisher","DOI":"10.1088\/1361-6463\/acbec3"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.1145\/3447818.3460371"},{"key":"ref42","article-title":"Sophisticated tools for performance analysis and auto-tuning of performance portable parallel programming","author":"Boehme"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.1109\/TC.2020.2964767"},{"key":"ref44","doi-asserted-by":"publisher","DOI":"10.1145\/3588993.3597262"},{"key":"ref45","doi-asserted-by":"publisher","DOI":"10.1145\/3437801.3441621"},{"key":"ref46","article-title":"Prior-guided Bayesian optimization","author":"Souza","year":"2020"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.1145\/3453483.3454109"},{"key":"ref48","doi-asserted-by":"publisher","DOI":"10.1109\/PMBS54543.2021.00017"}],"container-title":["IEEE Transactions on Parallel and Distributed Systems"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/71\/11083525\/11077458.pdf?arnumber=11077458","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,8,11]],"date-time":"2025-08-11T17:44:52Z","timestamp":1754934292000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11077458\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,9]]},"references-count":48,"journal-issue":{"issue":"9"},"URL":"https:\/\/doi.org\/10.1109\/tpds.2025.3587673","relation":{},"ISSN":["1045-9219","1558-2183","2161-9883"],"issn-type":[{"value":"1045-9219","type":"print"},{"value":"1558-2183","type":"electronic"},{"value":"2161-9883","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,9]]}}}